llm.ing

Design Arena — codetovideo

mirrorrun by Design Arenaunit: eloobserved Nov 19, 202625 modelssource ↗

Caveat: Head-to-head human votes on generated designs/code, mirrored via OpenRouter.

websitecodecategoriesgamedev3duicomponentdatavizsvgasciiartmobileappswebapps+19 more
#ModelScore
1Claude Fable 5.2
1456
2Claude Opus 5.5
1342
3Claude Sonnet 5.5
1320
4Kimi K3
1278
5Gemini 4 Carbon
1275
6Claude Opus 5
1249
7Gemini 4 Argon
1248
8Kimi K3.1 Pro
1242
9MiniMax-M3
1207
10Muse Spark 1.3
1204
11GPT-6.1 Sol
1194
12Claude Opus 4.8
1192
12Gemini 3.8 Flash
1192
14Gemini 3.7 Flash
1183
15Gemini 3.5 Flash
1162
16Grok 4.6
1155
17Grok 4.5
1125
18Gemini 3.1 Pro Preview
1123
19Claude Opus 4.6
1122
20Gemini 3.6 Flash
1118
21Claude Sonnet 5
1113
22Muse Spark 1.2
1109
23Grok 4.7
1107
24Muse Spark 1.1
1090
25Grok 4.3
881

Dot = elo on a 852–1485 scale (position, not length). Scores link to model pages; every number keeps its source on the model page.